Accelerating 5G networks with the digital twin and thread
May 20, 2020
May 20, 2020
You’ve probably heard about the promise of 5G mobile networks. They deliver content faster and serve a greater number of people than their predecessors, powering continually connected mobile experiences with better-than-WiFi quality.
But building a complete 5G network is a complex challenge. It’s not as simple as upgrading existing cell phone sites. Instead of today’s spaced-out towers, 5G requires a dense network of small cells positioned closed to the ground. That creates a complicated planning process for providers. They can’t rely on the approaches they used for planning previous networks, and they have a different supply chain for materials as well.
We’re developing technology solutions that let companies simulate the results of network planning at the earliest stages. Providers can test a proposed 5G network rollout without building a single cell or even visiting the site. How? With the digital twin and thread.
We’ve talked about the digital twin and thread concepts in a previous post on this blog. Our latest research shows that the twin and thread can change the way a 5G network operator does business, from the earliest stages of network planning all the way through to operation.
Predict better: network simulation
5G operators have a lengthy wait between initial network design and seeing revenue from their roll-out. Just choosing which sites to build out requires an understanding of customers, competition, and how the sites will perform. All those factors combine to determine the company’s return on capital investment, and gathering that information takes time. So we used digital environment twins to unify customer, competitor, and performance data quickly in a tool called the Network Decision Platform.
The platform takes data about population segments, demographics, and business firmographics at the census tract level and presents geographic visualizations and analysis. We also add visualizations of the current competitive landscape, broken down by type and number of offerings in each area. This information combines to helps planners determine which market has the most potential and choose where to launch.
But companies also need to consider what it costs to build a network in that market. For this challenge, we built an environment twin that incorporates 3D data of everything in the market’s ZIP code, from terrain to streets, buildings, and trees. Accurate to the cubic meter, our twin contains information about the material properties of the objects within it as well. From there, we can accurately simulate a 5G cell’s shorter waves to see what kind of service they’ll provide. Now network planners can see where to place new cell sites to create the best network.
Design smarter: Intelligent Site Builder
Once companies select a site to build, their engineering team determines the materials and parts necessary to complete the work. Engineers depend on information that may be incomplete or out of date, like item availability or reliability data. Without accurate information, creating the necessary the bill of materials can involve many iterations of material selection, auditing, and approval. This can delay the final build-out of the site and incur extra costs down the road.
Our Intelligent Site Builder (ISB) addresses these challenges by creating a digital site twin. It discovers cell sites similar to the target site, using their profile configurations to create a digital twin for new site deployment or upgrades. The tool infers the best materials for use and recommends parts to include. It offers suggestions for everything from major elements like radios down to minor parts like screws and fasteners, and learns over time. For example, if certain parts are always used together in installations, the ISB factors that into future suggestions. The tool also suggests alternatives if parts are out of stock, have long lead times, or have outstanding maintenance issues. Engineers can audit and select materials, approve recommendations, and inspect alternatives. ISB’s AI guidance reduces approval iterations and man-hours as well as the overall operations cost.
The ISB also offers unique insight into the supply chain delivery schedule and material demands. By mapping selected parts against the project’s construction schedule and milestones, the tool can identify materials that will arrive after construction start, for example. The engineering team can then select an alternate part or update the overall schedule to avoid delays.
The ISB can do all of this thanks to the power of a knowledge graph. (Click here for a previous post about the value of knowledge graphs in powering AI solutions.) The knowledge graph tracks connections between major and minor material parts while also capturing the evolving domain knowledge of thousands of site engineers and supply chain operators.
Operate smoother: virtual site visits
With a bill of materials created, the company can send a team out to build the site and bring it online. Ideally, a site build would happen all in one visit. In practice, though, sites don’t always line up with expectations. Access to power might be a few feet further away than expected, or the available area smaller than planned. When these kinds of problems happen, crews end up having to revisit the site to finish the installation, leading to more delays.
Once again, though, the digital twin offers a solution. Our previous environment twin for network planning needed a wide scale model for fast simulation. Now we need a close-up site twin, created using our “photogrammetry-as-a-service" pipeline: creating a digital representation of the site from imagery. With this twin, the build team can use a simple app to make sure that their plans for the site will work, like checking if a length of cable will comfortably reach between the power grid and the cell equipment. Using this twin, we can cut down the number of site visits by the build team, keeping the process moving quickly. And the pipeline continues after the build is complete, recording the state of the finished site so that the twin can be consulted for future maintenance or upgrades.
A digital thread connects information used by multiple parts of the business in a seamless and invisible way. That’s how our Intelligent Site Builder can incorporate information from procurement and inventory management without heavy effort from the engineering team. The thread ties these lines of business together without forcing an engineer to learn about procurement or forcing a planner to learn about part selection. It makes its powerful insights actionable for everyone.
But what’s more, the thread supports future learning. Information gleaned from virtual site visits, for example, informs the next iteration of network planning. In a similar way, as we learn what sites make for easier deployments, the Network Decision Platform’s definition of a potential site changes to accommodate the real-world experience. Every new action and piece of data helps create better recommendations for the future.
We’re continuing to explore the ways that the digital twin and thread can enhance security, maintenance, supply chain, and lifecycle decisions for 5G providers. And of course, the twin and thread offer similar value for companies across industries. Want to learn more about a specific application of the twin and thread for your client or industry? Read our PoV here, or contact Teresa Tung. For more information about the Network Decision Platform or Intelligent Site Builder, contact Colin Puri and Robert Dooley.